Extending a model of shift-work tolerance

The present study contributes to theory and practice through the development of a model of shift-work tolerance with the potential to indicate interventions that reduce nurses' intention toward turnover and increase job satisfaction in hospital-based settings. Survey data from 1257 nurses were used to conduct structural equation modeling that examine the direct and indirect effects of supervisor and colleague support, team identity, team climate, and control over working environment on time-based work/life conflict, psychological well-being, physical symptoms, job satisfaction, and turnover intention. The analysis of the proposed model revealed a good fit The chi-square difference test was non-significant (χ2(26)=338.56), the fit indices were high (CFI=.923, NFI=.918, and NNFI=.868), the distribution of residuals was symmetric and approached zero, the average standardized residual was low (AASR=.04), and the standardized RMR was .072. In terms of the predictor variable, the final model explained 48% of the variance in turnover intention. The data revealed considerable evidence of both direct effects on adjustment and complex indirect links between levels of adjustment and work-related social support, team identity, team climate, and control. Nurses with high supervisor and coworker support experienced more positive team climates, identified more strongly with their team, and increased their perceptions of control over their work environment. This in turn lowered their appraisals of their time-based work/life conflict, which consequently increased their psychological well-being and job satisfaction and reduced their physical health symptoms and turnover intention. The type of shift schedule worked by the nurses influenced levels of turnover intention, control over work environment, time-based work/life conflict, and physical symptoms.

The relationship between diurnal type and time duration estimation at morning and evening times of day

This study investigates whether different diurnal types (morning versus evening) differ in their estimation of time duration at different times of the day. Given that the performance of morning and evening types is typically best at their preferred times of day, and assuming different diurnal trends in subjective alertness (arousal?) for morning and evening types, and adopting the attentional gate model of time duration estimation, it was predicted that morning types would tend to underestimate and be more accurate in the morning compared to evening types where the opposite pattern was expected. Nineteen morning types, 18 evening types and 18 intermediate types were drawn from a large sample (N=1175) of undergraduates administered the Early/Late Preference Scale. Groups performed a time duration estimation task using the production method for estimating 20-s unfilled intervals at two times of day: 0800/1830. The median absolute error, median directional error and frequency of under- and overestimation were analysed using repeated-measures ANOVA. While all differences were statistically non-significant, the following trends were observed: morning types performed better than evening types; participants overestimated in the morning and underestimated in the evening; and participants were more accurate later in the day. It was concluded that the trends are inconsistent with a relationship between subjective alertness and time duration estimation but consistent with a possible relationship between time duration estimation and diurnal body temperature fluctuations. (C) 2002 Elsevier Ltd. All rights reserved.

Transcultural properties of the Composite Scale of Morningness: The relevance of the "morning affect" factor

Morningness scales have been translated into several languages, but it lack of normative data and methodological differences make cross-cultural comparisons difficult. This study examines the psychometric properties and factor structure of the Composite Scale of Morningness (CSM) in samples from five countries: France (n = 627), Italy (n, = 702), Spain (n = 391), Thailand (n. = 503), and Australia (17 = 654). Strong national differences are identified. A quadratic relationship between age and CSM total score was apparent in the Australian data with a downward trend after age 35 yrs. There was no age effect in air), sample in the range from 18 to 29 yrs. Factor analysis identified a three-factor solution in all groups for both men and women. Tucker's congruence coefficients indicate that: (1) this solution is highly congruent between sexes in each culture, and (2) a morning affect factor is highly congruent between cultures. These results indicate there are national differences in factorial structure and that cut-off scores used to categorize participants as morning- and evening-types should be established for different cultural and age groups.

Population-based continuous optimization, probabilistic modelling and mean shift

Evolutionary algorithms perform optimization using a population of sample solution points. An interesting development has been to view population-based optimization as the process of evolving an explicit, probabilistic model of the search space. This paper investigates a formal basis for continuous, population-based optimization in terms of a stochastic gradient descent on the Kullback-Leibler divergence between the model probability density and the objective function, represented as an unknown density of assumed form. This leads to an update rule that is related and compared with previous theoretical work, a continuous version of the population-based incremental learning algorithm, and the generalized mean shift clustering framework. Experimental results are presented that demonstrate the dynamics of the new algorithm on a set of simple test problems.